Signal optimization: using signals to mitigate information overload in crowdfunding campaigns

Originally designed to support artists and financially constrained businesses, crowdfunding evolved into a transformative alternative funding option for entrepreneurial ventures (Mollick, 2014). Rapid growth in crowdfunding is evident: in 2022 alone, campaigns raised over $64 billion, with projections suggesting this figure will exceed $75bn by 2027 (Statista, 2023). Despite these promising numbers, 85% of crowdfunding campaigns fail to meet their funding goals (Shepherd, 2023).

One of the challenges contributing to these failures is information overload (Moy et al., 2018; Thapa, 2020). Although the information provided by helps establish legitimacy and increases the likelihood of funding success (Liu et al., 2022; Fisher et al., 2017), excessive information can lead to overload, causing psychological fatigue and reducing the probability of success (Edmunds and Morris, 2000). Based on these arguments, past studies have provided evidence suggesting a curvilinear relationship, initially increasing and then decreasing, between the volume of information and crowdfunding success (Thapa, 2020).

The presence of a curvilinear relationship between information volume and funding success introduces practical challenges, especially for inexperienced fund-seekers who often present extensive information to establish credibility (Fisher et al., 2017). Therefore, should fund-seekers provide a large volume of information to increase their credibility? Given that the relationship between the volume of information and funding success may be curvilinear, addressing this question poses a significant challenge.

This prompts an important question: What tools assist fund-seekers in improving their probability of funding success, especially when providing a large volume of information? This study integrates insights from the information overload literature (Edmunds and Morris, 2000) with perspectives from signaling theory on signal portfolios and interactions (Kleinert, 2024; Di Pietro et al., 2023; Spence, 1978). It examines how fund-seekers on reward-based crowdfunding platforms can effectively manage information overload by designing an appropriate signal portfolio.

Building on recent studies applying signaling theory (Kleinert, 2024; Di Pietro et al., 2023), this research argues that the influence of signals, whether costly or costless, on funding success depends on the volume of information provided during the crowdfunding campaign. This study examines the context of information volume as low versus high (Thapa, 2020). A low volume of information refers to situations in which increasing information enhances funding success, whereas a high volume of information has the opposite effect (Thapa, 2020). This study identifies four distinct signals, two costly and two costless, and investigates their moderating influence on the curvilinear relationship between information volume and funding success.

We employed fixed-effects logit regression models to analyze data from 2,681 Kickstarter campaigns. The findings reveal an interesting pattern of how different signals moderate the curvilinear relationship between information volume and funding success. Specifically, operational transparency (a costless signal) and past crowdfunding experience (a costly signal) reduce the likelihood of funding success under conditions of low information volume but enhance it under high information volume. Conversely, perceived project authenticity (a costless signal) and perceived product quality (a costly signal) exhibit the opposite effect: they increase the probability of success with low information volume, but diminish with high. The findings indicate that the influence of signals, whether costly or costless, on funding success depends on the volume of information disclosed during a crowdfunding campaign.

This study makes several significant contributions to the literature. First, it advances the application of signaling theory in crowdfunding research by exploring signal interactions, an area that needs further investigation (Kleinert, 2024; Di Pietro et al., 2023; Tajvarpour and Pujari, 2022; Pollack et al., 2021). It also enhances our understanding of how signals interact in the crowdfunding context (Kleinert, 2024; Drover et al., 2018), thereby expanding the theoretical foundations of the field. Additionally, the results revealed the multifaceted nature of the signals and suggested opportunities for future research.

Second, this study contributes to discussions on information overload and psychological fatigue in crowdfunding (Mullins and Sabherwal, 2022; Thapa, 2020; Moy et al., 2018). While the existing literature often describes an inverse U-shaped relationship between information volume and crowdfunding success (Thapa, 2020; Moy et al., 2018), it has not thoroughly addressed factors that enhance funding success under conditions of high information volume. This study fills this gap by identifying the factors that improve the probability of success under such conditions.

Third, the findings offer practical guidance for fund-seekers by helping them devise strategies to enhance their likelihood of securing funding, especially when presenting a large volume of information. Traditionally, a high volume of information has been linked to reduced funding success (Thapa, 2020), and research has often overlooked strategies to assist fund-seekers in such scenarios (Moy et al., 2018; Thapa, 2020). Thus, this study provides valuable insights for fund-seekers when designing effective crowdfunding campaigns.

Crowdfunding is a method of raising funds from a large number of people, typically via the Internet, to support a project or venture (Belleflamme et al., 2014; Mollick, 2014). This involves aggregating numerous small contributions to support projects, ventures, or causes that might face challenges in accessing traditional financing avenues (Belleflamme et al., 2014; Mollick, 2014; Zhao et al., 2019). While crowdfunding has historical roots dating back to the 18th-century Irish Loan Funds and the 19th-century funding of the Statue of Liberty’s pedestal by New York citizens (Fundable, 2023; Startups Team, 2018; Zhao et al., 2019), its contemporary form emerged with ArtistShare, introducing reward-based crowdfunding for musicians in 2001 (Fundable, 2023; Startups Team, 2018; Zhao et al., 2019).

Several events have led to the expansion of crowdfunding, including the financial upheaval of 2008 and 2009, which led to the birth of platforms such as Indiegogo and Kickstarter (Fundable, 2023; Startups Team, 2018; Zhao et al., 2019). Another incident that formalized the existence of crowdfunding platforms was the JOBS Act. The JOBS Act paved the way for modern crowdfunding platforms, with a pivotal moment occurring in 2012, when Title III of the JOBS Act expanded equity investment to non-accredited investors (Zhao et al., 2019). In 2022, North American crowdfunding campaigns raised a whopping $17.2bn, a figure that is projected to continue to rise in the future (Shepherd, 2023). In an online crowdfunding campaign, the fund-seekers provide potential backers with valuable information about themselves, the project, and how the funds will be used in order to persuade them to support the project (Bi et al., 2017; Zacharakis and Shepherd, 2016).

The information provided by fund-seekers allows potential fund-providers to analyze a campaign and make well-informed decisions (Belleflamme et al., 2014; Mollick, 2014; Zacharakis and Shepherd, 2016). Well-organized information demonstrating credibility and viability enhances the decision-making process and increases the probability of funding success (Chen, 2023). Comprehensive and detailed project descriptions, along with transparent disclosure of objectives, methods, and expected outcomes, enhance backers' understanding of the project and increase the legitimacy of the project (Fisher et al., 2017). Additionally, increased clarity facilitates stronger connections between project creators and potential supporters. Therefore, fund-seekers are often motivated to provide large volumes of information to potential fund-providers (Moy et al., 2018; Chan et al., 2020).

An intriguing perspective arises when we consider the possibility of excess information being provided by the fund-seekers to potential fund-providers. A large volume of information requires the potential fund-providers to examine a large volume of information and decide whether to trust that information or not which potentially leads to decision fatigue (Hirshleifer et al., 2019). Furthermore, excessive information can result in cognitive overload, impeding information processing (Jackson and Farzaneh, 2012; Edmunds and Morris, 2000). Consequently, individuals may experience heightened anxiety, a sense of being overwhelmed and psychological/decision fatigue including impaired decision-making and impulsivity (Misra et al., 2020; Hirshleifer et al., 2019). When potential fund-seekers are confronted with a large volume of information, they either detach themselves from the project or from the internet for a break (Sonnentag, 2011). As a result, under high-information conditions, potential fund providers avoid investing in these projects, lowering the likelihood of funding success. Based on these arguments, we generate the following hypothesis:

Previous research has investigated this hypothesis, and the current study extends this body of work by proposing four additional hypotheses. Hypothesis 1 not only replicates the findings of prior studies but also serves as the foundation for hypotheses 2 through 5. Accordingly, Hypothesis 1 is developed as the starting point for our analysis.

This paper argued that, all else being equal, an initial increase in the volume of information improves the probability of funding success. However, if the volume of information continues to increase beyond an optimum point, it ultimately reduces the probability. However, crowdfunding campaigns, particularly those initiated by inexperienced creators with innovative ideas, must establish credibility and rapport with potential funders (Fisher et al., 2017). This requirement might compel fund-seekers to furnish a considerable amount of information. Thus, in certain circumstances, providing a large volume of information becomes a necessity. Therefore, to identify variables that moderate the curvilinear relationship this paper draws on signaling theory.

Multiple studies have applied signaling theory to identify the factors influencing funding success in crowdfunding campaigns. Signaling theory explains how people, organizations, or entities communicate and convey information to others in situations where there is an imbalance of knowledge (Spence, 1978). Signaling theory posits that actions can function as signals of hidden attributes, yielding benefits for both the signaler and the observer while maintaining the credibility of the signal (Spence, 2002).

Signaling theory has been applied to a wide range of social domains, including entrepreneurship, human resource management, and crowdfunding (Connelly et al., 2011). In the context of crowdfunding, signaling theory can help entrepreneurs and project creators convey their credibility and quality to potential backers when complete information is lacking (Kleinert, 2024). Entrepreneurs commonly employ signals, such as their track record of success, industry expertise, or endorsements from reputable sources, to communicate the viability and worthiness of their projects for support (Courtney et al., 2017; Ko and McKelvie, 2018). These signals primarily aim to reduce uncertainty among potential backers, ultimately influencing their decisions to contribute funds to crowdfunding campaigns (Fisher et al., 2017).

Signaling theory can be a valuable tool for entrepreneurs and project creators who are looking to raise funds through crowdfunding (Kleinert, 2024). By understanding how signaling theory works, fund-seekers can develop strategies to communicate their credibility and quality to potential backers, and ultimately increase their chances of success (Fisher et al., 2017). The traditional view of signaling theory has typically relied on costly signals (Kleinert, 2024; Connelly et al., 2011). However, recent studies have shown that both costly and costless signals can either directly or indirectly influence crowdfunding success (Di Pietro et al., 2023), and researchers should examine the collective effect of a signal portfolio on crowdfunding success (Kleinert, 2024; Di Pietro et al., 2023).

Extant literature often classifies signals as costly or costless. Costly signals require resources and investments that demonstrate commitment and confidence, and costless signals rely on rhetoric and messages to indicate intentions and ambitions (Kleinert, 2024; Di Pietro et al., 2023). The traditional view argues that costly signals are more effective (Connelly et al., 2011); however, recent researchers demonstrate that costless signals can also be effective, especially under special conditions (Kleinert, 2024; Di Pietro et al., 2023). Two recent perspectives, the interactionist perspective (Di Pietro et al., 2023; Pollack et al., 2021) and the portfolio perspective (Kleinert, 2024), provide a balanced perspective on how signals can influence crowdfunding success.

The interactionist perspective on signaling theory focuses on how costly and costless signals work together to influence crowdfunding success (Di Pietro et al., 2023). While costly signals demonstrate the creator’s commitment and competence, costless signals, such as regular communication and community engagement, build trust and authenticity. By effectively utilizing both types of signals, project creators can enhance their chances of attracting backers and achieving crowdfunding goals. The portfolio perspective on signaling theory argues that signals do not occur in isolation and researchers should focus on examining the collective influence of signals (Kleinert, 2024). Drawing on insights from interactionist and portfolio perspectives (Di Pietro et al., 2023; Pollack et al., 2021) this study argues that both costly and costless signals can be used to mitigate the effects of information overload in crowdfunding campaigns.

Researchers have illustrated that operational transparency, a costless signal (Kattwinkel and Knoepfle, 2023), can influence crowdfunding success (Mejia et al., 2019). Operational transparency pertains to the systematic practice of openly sharing information, procedures, and activities associated with the inner mechanisms of an organization, project, or undertaking (Kattwinkel and Knoepfle, 2023). In the context of crowdfunding research, operational transparency encompasses the degree to which individuals seeking funding divulge pertinent information regarding the operational facets of their project subsequent to its initiation, primarily achieved through the consistent provision of updates (Mejia et al., 2019). This form of operational transparency serves as a costless signal in crowdfunding campaigns, setting it apart from actions that require much money or sacrifices (Mejia et al., 2019). Updates leverage readily available information resources to underscore commitment and trustworthiness (Mejia et al., 2019). Therefore, operational transparency leads potential fund providers to perceive fund-seekers as possessing greater authenticity, expertise, and dedication, resulting in an increased probability of funding success (Mejia et al., 2019).

In the context of large volume of information, operational transparency bolsters credibility by reaffirming fund-seekers' dedication to fulfilling their promises (Buell et al., 2017; Buell and Norton, 2011). Operational transparency thus becomes a potent mechanism, mitigating information asymmetry, fostering confidence among backers, and amplifying the impact of comprehensive information provision on funding success (Buell et al., 2017; Buell and Norton, 2011). Information asymmetry occurs when one party in a transaction or interaction has more or better information than the other, leading to an imbalance in decision-making power (Bergh et al., 2019). Furthermore, backers tend to place their trust in campaigns that offer clear insights into operations, project management, and risk mitigation strategies (Mejia et al., 2019; Kattwinkel and Knoepfle, 2023). Building upon this premise, we formulate the following hypothesis:

In other words, operational transparency moderates the curvilinear relationship between information volume and the likelihood of funding success by influencing the shape and peak of the inverted-U relationship. Specifically, under high information volume conditions, operational transparency amplifies the positive effects of additional information, making the curve more pronounced and potentially shifting the optimal information amount.

To comprehensively evaluate the interaction effects of the costless signals (Mejia et al., 2019), we introduced an additional moderator into our model: perceived project authenticity. Past research has identified authenticity as a costless signal that is likely to impact the success of crowdfunding campaigns (Mejia et al., 2019; Oo and Allison, 2024). Authenticity can be approached from two distinct perspectives (Oo and Allison, 2024). First, it can be perceived as the extent to which observers believe emotional expressions are genuine (Ashforth and Tomiuk, 2000). Second, it can be defined by the alignment between one’s expressions and inner emotions (Grandey, 2003). This study aligns with Ashforth and Tomiuk’s (2000) perspective because, in social contexts involving interactions between two parties, the observer’s viewpoint carries significant weight (Oo and Allison, 2024). Additionally, this perspective contrasts with operational transparency, as it focuses on how potential fund providers interpret information rather than the actions and disclosures of fund-seekers (Radoynovska and King, 2019), underscoring the importance of including this construct in our model.

Perceived project authenticity refers to an individual’s subjective assessments and beliefs regarding a project’s genuineness, honesty, and integrity (Ashforth and Tomiuk, 2000; Oo and Allison, 2024). In crowdfunding campaigns, it signifies potential backers' subjective evaluations of how well a project’s representations, claims, and communications align with its actual intentions, actions, and overall credibility (Oo and Allison, 2024). Perceived project authenticity in crowdfunding campaigns is considered a costless signal because it relies on subjective assessments from fund providers and does not require substantial resource investments from fund-seekers (Oo and Allison, 2024). Prior studies have demonstrated that the perceived project authenticity of crowdfunding campaigns significantly influences the probability of funding success (Oo and Allison, 2024).

Under a large volume of information condition, perceived project authenticity can potentially become counterproductive for a crowdfunding campaign. In situations where a high amount of information is provided, potential fund providers may interpret a high level of authenticity as an attempt to compensate for potential credibility issues (Weischer et al., 2013). This paradoxically raises doubts about the project’s authenticity, prompting backers to question why a seemingly genuine and well-conceived project feels the need to inundate them with excessive information, which in turn may cast doubts on the project’s true viability. Relying on this information, we propose the following hypothesis:

To be more precise, perceived project authenticity moderates the curvilinear relationship between information volume and the probability of funding success by altering the shape and peak of the inverted-U relationship. Specifically, under high information volume conditions, perceived project authenticity diminishes the positive effects of additional information, flattening the curve and potentially reducing the optimal information amount.

As previously mentioned, prior research has underscored the need for further investigation into signal portfolio and the interaction effects of signals (Kleinert, 2024; Di Pietro et al., 2023). Consequently, we integrated additional costly signals into our research model, with the first such signal being perceived product quality. Given that the creation of high-quality products often demands a significant commitment of resources, we argue that perceived product quality represents a costly signal (Frederiks et al., 2019). In the context of new products and ventures, product quality encompasses actor-independent attributes, including desirability, novelty, and economic viability (Davidsson, 2015; Frederiks et al., 2019).

Past research has consistently highlighted that perceived product quality serves as a pivotal signaling mechanism in the context of entrepreneurial ventures seeking external funding (Davidsson, 2015; Piva and Rossi-Lamastra, 2018; Shah and Thapa, 2023). In our study, we contend that in situations marked by high information volume, perceived product quality takes on a negative moderating role. Our study posits that extensive information can lead backers to interpret an emphasis on product quality as an attempt to compensate for potential inadequacies, paradoxically undermining backers' trust (Kim et al., 2016). Furthermore, when perceived product quality is not consistently upheld throughout the wealth of information, backers may perceive inconsistencies that erode their confidence, ultimately adversely impacting funding success (Kim et al., 2016). Building on this argument, our study develops the following hypothesis:

To be more precise, perceived product quality moderates the curvilinear relationship between information volume and the probability of funding success by altering the shape and peak of the inverted-U relationship. Specifically, under high information volume conditions, perceived product quality reduces the positive effects of additional information, flattening the curve and potentially lowering the optimal information amount.

Past crowdfunding experience serves as a costly signal because of the resources, time, and effort required to launch and operate previous campaigns. Fund-seekers who possess experience in launching crowdfunding campaigns demonstrate a heightened level of competence, strategy, and understanding of the crowdfunding landscape (Kotha and George, 2012; Skirnevskiy et al., 2017). The study asserts that past crowdfunding experience, acting as a costly signal plays a significant moderating role in shaping the intricate relationship between the volume of information and the likelihood of funding success in crowdfunding campaigns (Skirnevskiy et al., 2017). This argument delves into the dynamic interplay between the fund-seeker’s experience and varying levels of information, elucidating how past crowdfunding experience influences backers' perceptions and decisions.

This paper argues that in situations characterized by high information volume, the past crowdfunding experience of the fund-seeker positively moderates the curvilinear relationship between information volume and funding success (Fortezza et al., 2021). As potential fund-providers navigate a surge of information, the fund-seeker’s experience assumes the role of a valuable indicator of their capacity to navigate campaign complexities and uphold commitments. Backers are more likely to interpret past successes as evidence of reliability, especially amidst the data deluge (Fortezza et al., 2021). This moderating effect amplifies the fund-seeker’s credibility, offsetting potential skepticism stemming from excessive information and ultimately contributing to heightened funding success. Based on these arguments, this study develops the following hypothesis:

To be more precise, past crowdfunding experience moderates the curvilinear relationship between information volume and the probability of funding success by shaping the steepness and peak of the inverted-U relationship. Specifically, under high information volume conditions, past crowdfunding experience enhances the positive effects of additional information, making the curve more pronounced and potentially increasing the optimal information amount.

The Figure 1 below presents the theoretical model for this study:

For this study, we utilized a dataset consisting of 3,221 Kickstarter campaigns initiated between 2013 and 2021. Similar datasets have been employed in prior studies (Geiger and Moore, 2022; Short et al., 2017). The researchers initially acquired campaign details including URLs, campaign names, campaign IDs, goals, and amount raised from kaggle.com. Subsequently, the researchers visited the URLs and cross-referenced campaign names and IDs, and collected relevant data. After listwise procedures were applied to eliminate incomplete data, the final dataset consisted of 2,681 campaigns.

The dependent variable for this study was crowdfunding success, defined as the achievement of a Kickstarter campaign’s funding goal, where the cumulative investment pledged by fund providers equals or exceeds the predetermined target set by the fund-seeker. While some studies have employed the total funds raised as a dependent variable (Mollick, 2014), the most accurate measure of campaign success, as supported by research (Hildebrand et al., 2017; Moritz and Block, 2016), is its ability to attain the funding goal. Therefore, we used the attainment of the funding goal as our primary measure of campaign success, aligning with the prevailing approach for assessing crowdfunding success (Courtney et al., 2017).

The primary predictor variable in this study was information volume, which included total text, videos, and images shared by fund-seekers with potential fund providers (Thapa, 2020). Total text encompassed the cumulative word count from the “my story” section, updates, and responses section; video length indicated the duration of videos uploaded by fund-seekers in seconds, and total images represented the number of images on the campaign page (Thapa, 2020). This data was collected from the campaign pages of Kickstarter projects accessible on the Kickstarter website, using the URLs obtained from the Kaggle website. Sum of total text, video length, and image count were used to determine the information volume (Baron and Tang, 2011; Toft-Kehler et al., 2016). The information volume was scaled to perform the regression analysis. This measurement approach aligns with the methods employed by Thapa (2020) and Kaminski and Hopp (2020).

In this study, perceived project authenticity refers to the genuine portrayal of a fund-seeker’s emotions in their crowdfunding project (Oo and Allison, 2024). We adapted and refined items from Oo and Allison (2024) to measure perceived project authenticity. These items were, (1) The description of the campaign seems fake (reverse coded), (2) the campaign creators appeared to be putting on an act in this campaign description (reverse coded), and (3) the campaign creators actually experienced the emotions they described in this campaign description. Each project received ratings from three individuals with prior experience in investing in crowdfunding campaigns, and we computed the average score for analysis purposes. The calculation of Cohen’s kappa facilitated using the “irr” package in the R program, yielded values ranging from 0.75 to 0.82.

In this research, “operational transparency,” a moderating variable, refers to the practice of openly and comprehensively disclosing various operational aspects, processes, and activities associated with a crowdfunding campaign to potential backers, contributors, and the general public. To assess operational transparency, we employed the method proposed by Mejia et al. (2019), which involves quantifying the number of work-related terms within the update section of campaigns. Drawing from Mejia et al. (2019), consider an update with the following content: “We have acquired essential supplies valued at $150,000”. Additionally, we have procured 35 power generators and skillfully loaded them onto five cargo containers.’ In this context, there are three terms related to operational tasks, encompassing ‘acquired,’ ‘procured,’ and ‘loaded.’”

Fund-seekers' crowdfunding experience was measured by quantifying the number of projects they initiated (Butticè et al., 2017). When a team launched a project, the measurement of the crowdfunding experience involved aggregating the total number of projects generated by all team members.

Perceived product quality was assessed using the consensual assessment technique, employing a rating scale ranging from 1 to 10 (Amabile, 1982). A panel of three evaluators, consisting of two angel investors and a subject matter expert, evaluated the product’s attributes, including desirability, novelty, and economic potential (Frederiks et al., 2019). The cumulative score from this assessment served as the measure of the product’s quality. Interrater reliability was computed using Cohen’s kappa, facilitated through the “irr” package in the R program, resulting in values ranging from 0.71 to 0.81. These results demonstrate robust interrater reliability (Hsu and Field, 2003). This approach has been employed in several prior studies (Frederiks et al., 2019; Shah and Thapa, 2023).

In this study, we compiled a list of control variables drawing upon previous research. The control variables include the funding goal amount (USD), duration, number of backers, readability, economic value orientation, social value orientation, tone positivity, and tone negativity. The funding goal in crowdfunding campaigns refers to the predetermined monetary target that a campaign creator aims to raise from backers and contributors within a specified timeframe. On platforms such as Kickstarter, a campaign must attain its funding goal to secure the pledged amount from potential fund providers.

Additionally, we controlled for the campaign duration. “Duration” in crowdfunding context refers to the specific period (in days) during which a campaign is active and open for contributions from backers. Projects on Kickstarter can have durations ranging from as short as 1 day to a maximum of 60 days. Kickstarter recommends a duration of 30 days or less, as their findings reveal that shorter campaigns tend to enjoy higher success rates and generate a beneficial sense of urgency for the project. Campaigns lasting over 60 days seldom succeed according to Kickstarter’s research (Kickstarter, 2023).

Chan et al. (2020) demonstrated a significant relationship between the readability of campaign text and the achievement of funding goals. Following their findings, we assess readability using the Flesch Reading Ease Score (FRES) and incorporate it as a controlled variable. Additionally, we employ the Hemingway App Score to measure text readability and include it as a control variable (Thapa, 2020). Moreover, Moss et al. (2018) exhibited a potential relationship between social value orientation and environmental value orientation on crowdfunding outcomes. Consequently, we incorporate these two variables as control variables in our analysis. Furthermore, since our primary research questions revolve around signals, we also controlled for tone positivity and tone negativity in the campaigns (Henry, 2008). A summary of the measurement of the variables is presented in the table below:

This study employs a rigorous analytical approach, utilizing descriptive statistics, logistic regressions with fixed effects, and predictive modeling to analyze sample characteristics and test research hypotheses. The dataset was first organized in an Excel spreadsheet, and all subsequent analyses were conducted using the R programming language. To handle missing data, we applied pairwise deletion, ensuring that the analysis retained a robust sample size of 2,681, which allowed for a thorough evaluation of the study’s objectives. First, we calculated the mean, standard deviation, and correlations. Next, we conducted logistic regression analysis with fixed effects to test the hypotheses. Finally, we report the model predictability results to assess the strength of each model. All statistical analysis was performed in R program.

Utilizing the R programming language, we computed the mean, standard deviation, and correlation statistics, which are presented in Table 1. The results from this analysis are presented in Table 2.

Table 2 presents the results of the descriptive and correlation analyses. The italicized correlation coefficients are statistically significant at a p-value < 0.05. The underlined numbers are standard errors of the correlation coefficients. The dependent variable, crowdfunding success, exhibited a statistically significant correlation with information volume. Additionally, the dependent variable also exhibited a statistically significant correlation with some moderating variables. This allows us to move ahead with the regression analysis.

We developed three logit regression models with fixed effects to perform hypothesis testing. Logit regression models are used to predict the probability of an outcome, such as whether a crowdfunding campaign succeeds or fails (Boateng and Abaye, 2019). The fixed effects help control for variables that do not change over time, ensuring that the analysis focuses on the impact of the key factors being studied (Borenstein et al., 2010). All three models controlled for category-fixed effects and year-fixed effects. The first regression model represents the control model in which the predictors are derived based on past studies. The equation below represents the first model:

where Funding_Success_it is the dependent variable, α_i represents category-fixed effects and δ_t represents year-fixed effects and U_it represents the error term. CV represents the collection of control variables. In the second model, we added the primary independent variable, i.e. volume of information and its square term. The second model helps to examine hypothesis 1. The third model analyzes the interaction effects hypothesized in hypotheses 2–5. As in model 1, model 2 and model 3 control for category fixed effects and year fixed effects. Table 3 provides the log-likelihood results from the regression analysis.

We assess model fit using the difference between null and residual deviance, AIC, Fisher score, and pseudo R-squared. As R-squared cannot be calculated for Logistic regression, we present pseudo-R-squared results in Table 3 (Allison, 2012). The null deviance represents deviance with only the intercept, while the residual deviance accounts for all predictors. A larger difference between null and residual deviance indicates better model fit; among the three models, Model 3 (difference = 3027.9) outperforms Model 1 (difference = 1304.2) and Model 2 (difference = 1317.8), indicating its superior fit. Moreover, Model 3 has the lowest AIC, signifying its better fit compared to the other models. The Fisher score iterations of 9 for Model 3 and 25 for both Models 1 and 2 further support Model 3’s superior fit. Table 3 includes coefficients (log-odds) along with their standard errors. Additionally, Table 4 presents confidence intervals corresponding to the results in Table 3.

We conducted a meticulous examination of multicollinearity within our research, recognizing its potential to impact the precision of effect size estimations. Power functions, including curvilinear and interaction terms, inherently introduce multicollinearity into models (Allison, 2012; Cortina, 1993). While some degree of multicollinearity is considered admissible in such cases (Allison, 2012; Cortina, 1993), we followed a systematic approach to assess and address it. As part of this examination, we mean-centered the variables and computed VIFs for models that included both interaction terms and the linear term of information volume. The VIF values remained below the conservative threshold of 5.0 0 (Hair, 2011). Specifically, we observed VIF values of 4.31 for complex words, 3.22 for product quality, and 3.92 for the interaction between total information and product quality. All remaining VIF values were below 2.0, with the mean VIF across variables computed at 1.81. These results indicate no significant evidence of multicollinearity. Thus, we conclude that multicollinearity does not compromise the robustness or reliability of our findings.

Table 3 presents results from three fixed-effects logit regression models. We chose the fixed-effects logit regression model to account for category and year influences. Model 1 serves as the baseline, Model 2 includes focal predictors, and Model 3 represents the full interaction model. AIC, pseudo R-squared, residual deviance and Fisher score collectively indicate that Model 3 offers a superior fit compared to Models 1 and 2.

The findings from the fixed effects logit regression analysis, as displayed in Table 3, indicate a robust and statistically significant positive relationship between the volume of information (β = 7.043***, S.E. = 1.127, p < 0.001) and the probability of attaining funding success. Furthermore, the square of the information volume exhibited a statistically significant negative influence (β = −4.527***, S.E. = 0.710, p < 0.001) on the likelihood of achieving funding success. Collectively, these two findings suggest an inverse U-shaped relationship between the volume of information provided by the fund-seeker and the probability of funding success. These results remain consistent across model 2 and model 3, which provides strong empirical support for hypothesis 1. Model 3, the full model, was used to calculate the optimal level of information volume using the formula X* = −β1/(2β2), yielding an optimal level of 0.78. As the estimates are standardized (mean-centered), this suggests that the optimal level occurs at 0.78 standard deviations above the mean of the information volume, assuming all other factors remain constant. However, the moderating variables in the model indicate that this relationship may vary depending on the specific values of these moderating factors. These results above indicate an inherent liability for campaigns that require sharing a large volume of information which highlights the importance of this study in examining how certain costly and costless signals moderate these relationships.

We introduced operational transparency, a costless signal, as a moderator in the third model and the results suggest that under a large volume of information condition, operational transparency significantly and positively (β = 1.127***, S.E. = 0.322, p < 0.001) moderates the curvilinear relationship between information volume and probability of crowdfunding success. Another costless signal introduced in the third model was perceived project authenticity. The results indicate that, under a large volume of information condition, perceived project authenticity significantly and negatively (β = −1.318**, S.E. = 0.422, p < 0.01) moderates the relationship between information volume and probability of funding success. The results suggest that when crowdfunders have a large volume of information to share, they should prioritize maximizing operational transparency. Additionally, it is crucial for them to recognize that overly emphasizing project authenticity might diminish the probability of funding success. This results provides empirical support for hypothesis 2 and 3.

In addition to the above two costless signals, our full model also introduced two costly signals, i.e. perceived product quality and crowdfunding success. The results indicate that perceived product quality significantly and negatively (β = −1.027***, S.E. = 0.261, p < 0.001) moderates the relationship between information volume squared and the probability of crowdfunding success. Finally, crowdfunding experience, a costless signal, positively moderated the effect of information volume squared (β = 7.085***, S.E. = 0.547, p < 0.001) on crowdfunding success. These findings suggest that when faced with information overload, fund-seekers should emphasize their past experience. Additionally, if they already possess a significant amount of information to share, they should exercise caution in overemphasizing the quality of their product. These results provide support for hypotheses 4 and 5.

A deeper examination of the moderating variables reveals compelling insights. Operational transparency and crowdfunding experience demonstrated positive moderation in high information volume situations, but their effects turned negative in low information volume scenarios. Similarly, perceived project authenticity and perceived product quality exhibited negative moderation under high information volume, contrasting with their positive effects in low information volume conditions. These findings highlight the importance for fund-seekers to exercise caution when constructing their signal portfolio to manage information overload effectively.

Apart from scrutinizing the precise direction and extent of the relationship between the predictors and the outcome variable, we conducted an assessment of the predictive capacity inherent in each model. We maintain that our meticulous selection of predictors, independent variables, moderating variables, and interaction effects is pivotal, as it significantly augments the model’s predictive capability. The detailed presentation of the predictive capacity of all three models is outlined in Table 5 below:

The table above illustrates that Model 1 achieved a 68.32% accuracy in predicting crowdfunding successes and a 94.8% accuracy in predicting failures, with an overall average successful prediction rate of 85.57%. Model 2 exhibited a slight improvement in the predictive capacity of successful campaigns, although the accuracy of failure prediction decreased. The overall prediction success of Model 2 was only marginally superior to that of Model 1. However, substantial enhancements were observed with Model 3. It demonstrated an overall prediction accuracy of 98.32%. Specifically, the success prediction was accurate 97.57% of the time, and the failure prediction was correct 98.73% of the time. Consequently, the results indicate that Model 3 achieved an exceptionally high rate of successful predictions.

Recent research highlights a challenge for crowdfunding campaigns: managing information overload (Moy et al., 2018; Thapa, 2020). While widely recognized that information shared by fund-seekers enhances funding prospects, certain studies suggest that excessive information can diminish the probability of funding success (Moy et al., 2018; Thapa, 2020). This study employs signaling theory (Spence, 1978; Connelly et al., 2011) and the recent interactionist perspective on signal interaction (Kleinert, 2024; Di Pietro et al., 2023; Pollack et al., 2021) to assist fund-seekers in managing information overload.

Extant literature suggests a curvilinear relationship between information volume and crowdfunding success, indicating an optimal information volume that maximizes funding success (Moy et al., 2018; Thapa, 2020), and Hypothesis 1 replicates these findings while serving as the foundation for hypotheses 2–5 in our analysis. In this study, information below this optimum is termed low volume, while information exceeding it is labeled high volume. Building on prior research, we identify four signals—two costly and two costless—that moderate the curvilinear relationship between information volume and funding success and evaluate the moderating effect under both low and high information volume conditions. Operational transparency and perceived project authenticity, the two costless signals, exhibited opposing moderating influences in the curvilinear relationship between information volume and the probability of funding success. Similarly, perceived product quality and prior crowdfunding experience, the two costly signals, also showed opposite moderating influences.

The results revealed that operational transparency, a costless signal, and prior crowdfunding experience, a costly signal, had a negative moderating effect under low information volume and a positive effect under high information volume. Similarly, perceived project authenticity, a costless signal, and perceived product quality, a costly signal, had a positive moderating effect under low information volume and a negative effect under high information volume.

These results emphasize the intricate impact of signals on the relationship between information volume and funding success. Furthermore, they underscore the nuanced influences of signals in crowdfunding campaigns, providing valuable insights into how fund-seekers can strategically adjust their approaches based on the volume of information they provide to enhance their crowdfunding success.

This study adds to the discussion on whether signals need to be costly to be effective (Steigenberger and Wilhelm, 2018). Traditional signaling theory holds that signals must be costly to prevent inferior ventures from imitating them (Connelly et al., 2011), but modern versions relax this condition and assume that potential fund providers can back unverifiable claims by new ventures (Anglin et al., 2020). In nontraditional fundraising contexts, costless signals may still be required because the environments tend to feature noise and less sophisticated audiences (Anglin et al., 2018; Steigenberger and Wilhelm, 2018). Some researchers have highlighted the importance of credibility in noisy environments (Kleinert, 2024).

While credibility is an important signal, the results from this study show that whether potential funders reward or penalize signals depends on the nature of individual signals regardless of whether those signals are costly or costless. The study offers entrepreneurs the novel insight that signals, regardless of whether they are costly or costless, should be carefully tailored. The study argues for a signal portfolio perspective (Kleinert, 2024) that takes into account the effects of all signals by detailing how costly and costless signals work. This perspective acknowledges that individuals send multiple signals and that the effects of these signals can interact in complex ways. The study provides a framework for understanding how individuals can strategically manage their signal portfolio to maximize their chances of crowdfunding success.

Furthermore, this paper contributes to the emerging research on signal interactions (Kleinert, 2024; Di Pietro et al., 2023; Steingenberger and Wilhelm, 2018) by offering a deeper understanding of how costly and costless signals influence the relationship between the volume of information and the probability of crowdfunding success. More recently, researchers have examined novel approaches through which signals influence crowdfunding success (Anglin et al., 2020) and called on additional research on signal interaction (Kleinert, 2024; Di Pietro et al., 2023). This paper answers that call but differs from previous research by examining how costly and costless signals moderate the inverse U-shaped relationship between the volume of information and crowdfunding success. The results from this study reveal that individual signals, regardless of whether they are costly or costless, exhibit different moderating effects on the inverse-U-shaped relationship. Our results suggest that fund-seekers should examine the influence of individual signals, regardless of whether they choose costly or costless signals while managing information overload in crowdfunding campaigns.

Furthermore, the results from this study highlight the importance of creating signal portfolios in increasing the probability of crowdfunding success (Kleinert, 2024; Drover et al., 2018). Recent studies have attempted to compare the relative influence of costly and costless signals and exhibited that the effects of costly signals can be amplified by costless signals (Kleinert, 2024; Di Pietro et al., 2023; Colombo, 2021). We diversify this body of literature by exhibiting that both costly and costless signals can amplify the effects of other variables and increase the probability of funding success. While multiple signals can create portfolio inconsistencies (Bafera et al., 2020), crowdfunding campaigns often represent a noisy environment where fund-seekers disseminate a high volume of information (Thapa, 2020; Mollick, 2014), and a noisy environment often leads to dissemination of inconsistent signals (Steigenberger and Wilhelm, 2018). Based on the results from this study, we argue that examining the effect of each signal is important in a noisy environment to maximize the signal portfolio’s value.

This study also contributes to the literature on information overload in crowdfunding campaigns. The literature on information overload has largely described how information overload occurs and prescribed some ways to manage the problem of information overload (Edmunds and Morris, 2000). While past studies in the area of crowdfunding campaigns have highlighted that information overload can occur in crowdfunding campaigns (Moy et al., 2018; Thapa, 2020), few, if any, studies have presented plans to deal with the negative results of information overload. Considering that it is widely accepted that crowdfunding campaigns present a noisy environment (Steigenberger and Wilhelm, 2018), the importance of studying ways to deal with information overload is even higher. This study provides a signal portfolio perspective on dealing with information overload while designing a crowdfunding campaign. Therefore, this study broadens the application of information overload literature by combining it with the signal portfolio perspective of signaling theory.

This study applies signaling theory to offer actionable strategies for entrepreneurs and fund-seekers designing crowdfunding campaigns (Ahlers et al., 2015). It addresses the challenge of information overload by emphasizing the need to tailor signals thoughtfully, ensuring they are clear, relevant, and impactful (Kuratko and Audretsch, 2022; Yang et al., 2020; Thapa, 2020; Moy et al., 2018). Specifically, fund-seekers can enhance their campaigns by employing distinct signaling strategies tailored to the volume of information they provide.

Under conditions of high information volume, fund-seekers can leverage operational transparency to increase the likelihood of funding success. For instance, providing frequent and detailed updates on key milestones, such as securing resources, completing prototypes, or establishing partnerships, reassures backers of the campaign’s progress. Additionally, maintaining authenticity by presenting honest, relatable narratives and backstories can counteract potential skepticism associated with excessive information.

In contrast, under low information volume, fund-seekers should emphasize product quality and their past crowdfunding success. Highlighting product quality through endorsements, certifications, or demonstrable benefits can build confidence among backers. Similarly, showcasing past successes, such as high fulfillment rates or positive testimonials from previous campaigns, reinforces credibility and trustworthiness.

This study also underscores the importance of managing signal portfolios strategically. Treating each signal as a distinct yet complementary component of the overall campaign message (Kleinert, 2024) enables fund-seekers to craft consistent and compelling narratives. Such an approach is particularly vital in noisy crowdfunding environments, where strategic communication can mitigate the negative effects of information overload. By integrating these insights and strategies into their campaign designs, fund-seekers can enhance their credibility, communicate effectively, and maximize their chances of securing funding in the competitive crowdfunding landscape.

While this study offers valuable insights into the complex interplay between information volume, costly and costless signals, and crowdfunding success, it is not without its limitations. A key limitation of this research is its focus on reward-based crowdfunding campaigns, which restricts the broader applicability of the results to other forms of crowdfunding. Different crowdfunding models, such as equity-based or donation-based campaigns, may involve distinct audience dynamics, decision-making processes, and information-processing mechanisms (Ahlers et al., 2015). As a result, the findings of this study may not fully capture the nuances of how signals influence success in these other contexts, potentially limiting the transferability of the results across diverse fundraising settings.

Second limitation arises from the number of raters for assessing perceived project authenticity and product quality. In assessing perceived project authenticity and product quality, we utilized a panel of three raters, a common practice in similar research on crowdfunding (e.g. Frederiks et al., 2019; Shah and Thapa, 2023). While we recognize that using a larger number of raters could further enhance the reliability of the ratings, we opted for three raters based on practical considerations, such as resource constraints and statistical guidelines for inter-rater reliability. To ensure the robustness of our assessments, we calculated Cohen’s kappa, which showed strong agreement among the raters (kappa values ranging from 0.71 to 0.81). This level of agreement indicates a reliable assessment of the subjective attributes of authenticity and product quality. While future studies may benefit from a larger panel of raters, we believe the use of three raters strikes an appropriate balance between feasibility and reliability in the context of this study.

Another limitation of this study is related to the use of specific timeframe. As crowdfunding platforms evolve and new ones emerge, the dynamics of campaign success may shift over time, influenced by platform features, user behaviors, or changing trends in online interactions. The temporal and platform-specific nature of this study means that the observed relationships between signals and crowdfunding outcomes may not hold universally across all platforms or time periods (Stuart et al., 2015). Future research could strengthen the generalizability of these findings by examining campaigns on a variety of platforms, and over longer time horizons, to assess the robustness of the observed effects across different contexts and technological advancements.

Moreover, this study focuses primarily on the moderating role of signals but does not explore other potential mediating or moderating factors that may influence crowdfunding outcomes. Variables such as campaign type, creator experience, and audience engagement could play significant roles in shaping the effectiveness of different signals (Di Pietro et al., 2023; Guenther et al., 2018). A more nuanced investigation of these factors could provide a deeper understanding of the underlying mechanisms and broaden the scope of the study’s conclusions.

Additionally, the study does not account for external macro-level factors, such as economic conditions or political events, which could have a significant impact on crowdfunding campaigns. These factors may influence funders’ perceptions, investment behaviors, or trust in campaign creators, and thus affect the effectiveness of the signals identified in this study (Antonakis et al., 2010, 2014). Future research that integrates macroeconomic or political variables could offer a more comprehensive understanding of how signaling strategies interact with real-world conditions, enhancing the practical applicability of the findings to actual crowdfunding environments.

In terms of future research directions, several avenues warrant further exploration. First, the temporal aspects of signal effectiveness represent an important area for investigation (McQuail et al., 2021). It remains unclear how the influence of signals evolves throughout the course of a crowdfunding campaign. Understanding whether and how the salience of different signals changes over time could provide valuable insights into funders' decision-making processes and how they perceive campaign progress (Kaminski and Hopp, 2020).

Second, there is potential to explore the role of individual differences among funders. Research could investigate how factors such as risk tolerance, prior investment experience, or cultural background influence the interpretation and effectiveness of different signals. This would enrich our understanding of the heterogeneity among crowdfunding audiences and provide insights into how campaigns can tailor their signaling strategies to different demographic groups.

Third, future studies could explore the interplay between online and offline signals (Tosatto et al., 2022; Lanzolla and Frankort, 2016). With the rise of social media and online networking, understanding how signals manifest and interact across various platforms could provide a more holistic view of signaling dynamics in crowdfunding campaigns. Fourth, considering the evolving nature of crowdfunding and online communication, the research could focus on emerging forms of signals, such as multimedia presentations, virtual reality experiences, or interactive prototypes (Tosatto et al., 2022). Investigating the effectiveness of these novel signals in capturing funders' attention and trust could shed light on future trends in crowdfunding communication strategies. Finally, future research could examine how highlighting an offering’s sustainability benefits affects its potential funding success and how such sustainability-related information interacts with costly and costless signals (Lourenco et al., 2012).

While this study advances our understanding of signaling strategies in crowdfunding campaigns, there are still unexplored avenues and limitations to address. Future research efforts should aim to overcome these limitations, exploring diverse contexts and incorporating new variables to provide a comprehensive and nuanced understanding of signaling dynamics in the evolving landscape of crowdfunding.

This study investigates the intricacies of crowdfunding, with a particular focus on the delicate balance between the volume of information and the influence of signals on campaign success. By combining traditional signaling theory with a nuanced perspective on signal portfolio, interaction and information overload, this research provides valuable insights for both theory and practice. The study reaffirms the existence of an optimal information threshold and the crucial role of signals in crowdfunding success. Specifically, signals such as operational transparency and prior experience have a negative influence on funding success under low information volume, but a positive influence under high-information scenarios. Conversely, signals such as perceived project authenticity and product quality initially enhance success, but their impact diminishes as the volume of information increases. This research provides fundraisers with strategic guidance, offering a nuanced model for effective crowdfunding. While practical guidelines are valuable, further research exploring contextual factors and backers' psychological aspects is necessary to enhance our understanding of crowdfunding dynamics. Ultimately, this study empowers fundraisers with valuable tools to optimize their campaigns and launch a successful crowdfunding project.